Common inflammatory markers in the screening of knee arthroprosthesis infections.

Aim To evaluate the sensitivity and specificity of serum C-reactive protein (CRP) in early and late total knee arthroplasty (TKA) infections. Methods Blood tests to determine CRP levels (cut-off 10 mg/L)were conducted before surgery, at 1st day, 7th day and 15th day after surgery and at 1, 3, 6,12, 24 and 36 months. Patients had routine follow-up visits and radiological evaluations at 14 days and at 1, 3, 6, 12, 24 and 36 months. Infections were recorded and classified according to Widmer classification. The χ2 test or Fisher (in subgroups smaller than 10 patients) exact test was used to compare categorical variables. The statistical significance was set at p <0.05. Results A total of 19 infections were diagnosed during the followup. According to Widmer, five were classified as early post-operative and 14 as late chronic. All patients with early infections had suspected symptoms such as fever, swelling and pain. During the first month, 59 patients who had high CRP level but negative microbiological culture were considered as false positive representing a CRP sensitivity of 80% and a specificity of 67.6%. Fourteen patients had late chronic infection. Conclusion This study suggests that a synovial fluid aspiration should be performed in patients with persistent inflammation symptoms with or without radiographic signs of loosening. Moreover, it recommends the use of different serum and synovial tests for periprosthetic joint infection (PJI) diagnosis.